library(magrittr)
library(tidyverse)
library(Seurat)
library(readxl)
library(cowplot)
library(colorblindr)
library(viridis)
library(progeny)
library(destiny)

coi <- params$cell_type_super
cell_sort <- params$cell_sort
cell_type_major <- params$cell_type_major
louvain_resolution <- params$louvain_resolution
louvain_cluster <- params$louvain_cluster

1 Cluster markers

1.1 Major Fibroblast.super markers for cell assign

### load all data ---------------------------------
source("_src/global_vars.R")

# seu_obj <- read_rds(paste0("/work/shah/isabl_data_lake/analyses/16/52/1652/celltypes/", coi, "_processed.rds"))
seu_obj <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/outs_pre/", coi, "_seurat_", louvain_resolution, ".rds"))
# seu_obj <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_highqc.rds"))

myfeatures <- c("umapharmony_1", "umapharmony_2", "sample", louvain_cluster, "doublet", "nCount_RNA", "nFeature_RNA", "percent.mt", "doublet_score")

plot_data_wrapper <- function(cluster_res) {
  cluster_res <- enquo(cluster_res)
  as_tibble(FetchData(seu_obj, myfeatures)) %>% 
    left_join(meta_tbl, by = "sample") %>% 
    rename(cluster = !!cluster_res) %>% 
    mutate(cluster = as.character(cluster),
           tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite)))
}

plot_data <- plot_data_wrapper(louvain_cluster)

1.2 Subtype currated markers

helper_f <- function(x) ifelse(is.na(x), "", x)

markers_v7_super[[coi]] %>% 
  group_by(subtype) %>% 
  mutate(rank = row_number(gene)) %>% 
  spread(subtype, gene) %>% 
  mutate_all(.funs = helper_f) %>% 
  formattable::formattable()

rank	Activated.CAF.IGF1	Activated.CAF.ISG	Activated.CAF.TGFb	Angiogenic.CAF	Cycling.CAF	dissociated	Early.CAF.1	Early.CAF.2	Mesothelial.CAF.IL1	Pericyte
1	APOE	CXCL10	ACTA2	ANGPTL4	CDC20	IGKC	ACKR4	C7	AQP1	A2M
2	ATF3	IFI44L	COL11A1	BNIP3	CDK1	MALAT1	APOD	CADPS	C3	ACTA2
3	CYR61	IFIT1	COL12A1	CA12	MKI67	MTRNR2L12	C7	IGFBP5	CALB2	ADIRF
4	FBLN1	IFIT2	COL5A1	CA7	PTTG1	MTRNR2L8	CFD	LAMP5	CCDC80	CAV1
5	IGF1	IFIT3	COL5A2	EGLN3	TOP2A	SERPINF1	GSN	PEG3	CLDN1	CCDC102B
6		ISG15	COL6A1	ENO1			MGP	RBP1	EZR	COL4A1
7		MX1	COL6A3	ENO2			PEG3	RNASE1	HP	COL4A2
8		RSAD2	FAP	HILPDA					KRT18	CRIP1
9			FN1	LDHA					KRT19	HIGD1B
10			MMP11	VEGFA					KRT8	IGFBP7
11			MMP13						PRG4	MCAM
12									UPK3B	MEF2C
13										MHY11
14										NDUFA4L2
15										PP1R14A
16										RGS5

1.3 Subtype cluster markers

## define patient specific clusters
patient_specific_clusters <- plot_data %>% 
  group_by(patient_id_short, cluster) %>% 
  tally %>% 
  group_by(cluster) %>% 
  mutate(nrel = n / sum(n),
         ntotal = sum(n)) %>% 
  arrange(desc(nrel)) %>% 
  distinct(cluster, .keep_all = T) %>% 
  filter(nrel > 0.5) %>% 
  ungroup %>% 
  mutate(cluster_label_ps = make.names(paste0("OV.", patient_id_short, ".specific"), unique = T),
         cluster = as.numeric(cluster)) %>% 
  distinct(cluster, cluster_label_ps)

# marker_tbl <- read_tsv(paste0("/work/shah/isabl_data_lake/analyses/16/52/1652/celltypes/", coi, "_markers.tsv")) %>% 
#   filter(resolution == louvain_resolution)
marker_tbl <- read_tsv(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/outs_pre/", coi, "_markers_", louvain_resolution, ".tsv"))
# marker_tbl <- read_tsv(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_highqc_markers_02.tsv"))

## Hypergeometric test --------------------------------------

test_set <- marker_tbl %>% 
  group_by(cluster) %>% 
  filter(!str_detect(gene, "^RPS|^RPL")) %>% 
  slice(1:50) %>% 
  mutate(k = length(cluster)) %>% 
  ungroup %>%
  select(cluster, gene, k) %>% 
  mutate(join_helper = 1) %>% 
  group_by(cluster, join_helper, k) %>%
  nest(test_set = gene)

markers_doub_tbl <- markers_v7 %>% 
  enframe("subtype", "gene") %>% 
  filter(!(subtype %in% unique(c(coi, cell_type_major)))) %>% 
  unnest(gene) %>% 
  group_by(gene) %>% 
  filter(length(gene) == 1) %>% 
  mutate(subtype = paste0("doublet.", subtype)) %>% 
  bind_rows(tibble(subtype = "Mito.high", gene = grep("^MT-", rownames(seu_obj), value = T)))

ref_set <- markers_v7_super[[coi]] %>% 
  bind_rows(markers_doub_tbl) %>% 
  group_by(subtype) %>% 
  mutate(m = length(gene),
         n = length(rownames(seu_obj))-m,
         join_helper = 1) %>% 
  group_by(subtype, m, n, join_helper) %>%
  nest(ref_set = gene)

hyper_tbl <- test_set %>% 
  left_join(ref_set, by = "join_helper") %>% 
  group_by(cluster, subtype, m, n, k) %>%
  do(q = length(intersect(unlist(.$ref_set), unlist(.$test_set)))) %>%
  mutate(pval = 1-phyper(q = q, m = m, n = n, k = k)) %>%
  ungroup %>%
  mutate(qval = p.adjust(pval, "BH"),
         sig = qval < 0.01)

# hyper_tbl %>% 
#   group_by(subtype) %>% 
#   filter(any(qval < 0.01)) %>%
#   ggplot(aes(subtype, -log10(qval), fill = sig)) +
#   geom_bar(stat = "identity") +
#   facet_wrap(~cluster) +
#   coord_flip()
  
low_rank <- str_detect(unique(hyper_tbl$subtype), "doublet|dissociated|Mito.high|OV\\.[0-9]")
subtype_lvl <- c(sort(unique(hyper_tbl$subtype)[!low_rank]), sort(unique(hyper_tbl$subtype)[low_rank]))
  
cluster_label_tbl <- hyper_tbl %>% 
  mutate(subtype = ordered(subtype, levels = subtype_lvl)) %>% 
  arrange(qval, subtype) %>%
  group_by(cluster) %>% 
  slice(1) %>% 
  mutate(subtype = ifelse(sig, as.character(subtype), paste0("unknown_", cluster))) %>% 
  select(cluster, cluster_label = subtype) %>% 
  ungroup %>% 
  mutate(cluster_label = make.unique(cluster_label, sep = "_")) %>% 
  left_join(patient_specific_clusters, by = "cluster") %>% 
  mutate(cluster_label = ifelse(is.na(cluster_label_ps), cluster_label, cluster_label_ps)) %>% 
  select(-cluster_label_ps)

seu_obj$cluster_label <- unname(deframe(cluster_label_tbl)[as.character(unlist(seu_obj[[paste0("RNA_snn_res.", louvain_resolution)]]))])
plot_data$cluster_label <- seu_obj$cluster_label

cluster_n_tbl <- seu_obj$cluster_label %>% 
  table() %>% 
  enframe("cluster_label", "cluster_n") %>% 
  mutate(cluster_nrel = cluster_n/sum(cluster_n))

marker_sheet <- marker_tbl %>% 
  left_join(cluster_label_tbl, by = "cluster") %>% 
  group_by(cluster_label) %>% 
  filter(!str_detect(gene, "^RPS|^RPL")) %>% 
  slice(1:50) %>% 
  mutate(rank = row_number(-avg_logFC)) %>% 
  select(cluster_label, gene, rank) %>% 
  ungroup %>% 
  mutate(cluster_label = ordered(cluster_label, levels = unique(c(names(clrs$cluster_label[[coi]]), sort(cluster_label))))) %>% 
  spread(cluster_label, gene) %>% 
  mutate_all(.funs = helper_f)

marker_tbl_annotated <- marker_tbl %>% 
  left_join(cluster_label_tbl, by = "cluster") %>% 
  left_join(cluster_n_tbl, by = "cluster_label") %>% 
  select(-cluster, -resolution) %>% 
  mutate(cluster_label = ordered(cluster_label, levels = unique(c(names(clrs$cluster_label[[coi]]), sort(cluster_label))))) %>% 
  arrange(cluster_label, -avg_logFC, p_val_adj)

write_tsv(marker_sheet, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_marker_sheet.tsv"))

write_tsv(marker_sheet, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/supplementary_tables/", coi, "_marker_sheet.tsv"))

write_tsv(marker_tbl_annotated, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_marker_table_annotated.tsv"))

write_tsv(marker_tbl_annotated, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/supplementary_tables/", coi, "_marker_table_annotated.tsv"))

formattable::formattable(marker_sheet)

rank	Early.CAF.1	Early.CAF.2	Activated.CAF.ISG	Activated.CAF.TGFb	Activated.CAF.IGF1	Angiogenic.CAF	Mesothelial.CAF.IL1	Cycling.CAF	Pericyte	doublet.Endothelial.cell	doublet.Monocyte	doublet.Ovarian.cancer.cell	OV.068.specific	OV.075.specific	OV.105.specific	OV.105.specific.1
1	CFD	C7	CXCL10	MMP11	APOE	MT3	HP	CENPF	COL4A1	VWF	SPP1	MMP7	IGKC	MMP1	TIMP1	SLPI
2	APOD	RBP1	ISG15	CTHRC1	APOC1	SERPINE1	SLPI	H2AFZ	RGS5	FABP4	SRGN	TFPI2	FABP4	RARRES2	MMP9	WFDC2
3	C7	IGFBP5	IFIT1	COL11A1	IGF1	ANGPTL4	PRG4	TOP2A	COL4A2	A2M	CCL4	KRT7	COMP	DIO2	MT2A	CXCL1
4	MGP	RNASE1	IFIT3	POSTN	C3	PLIN2	KRT19	MKI67	PPP1R14A	PECAM1	C1QB	SPINT2	SFRP2	MMP11	APOC1	FDCSP
5	GSN	PEG3	RSAD2	CTSK	CCDC80	HILPDA	PLA2G2A	TUBA1B	IGFBP7	PLVAP	C1QA	LCN2	FGF7	TDO2	MIF	CXCL2
6	GPX3	CADPS	MX1	COL1A2	CXCL2	NDRG1	KRT18	STMN1	NDUFA4L2	CLDN5	TYROBP	CLDN3	RBP7	ISG15	ANGPTL4	PLAT
7	ADH1B	LAMP5	IFIT2	COL12A1	CXCL12	VEGFA	EZR	NUSAP1	MCAM	EGFL7	RGS1	EPCAM	CILP	TNFSF10	SEC61G	MT2A
8	TNXB	COLEC11	IFI44L	VCAN	FBLN1	IGFBP3	UPK3B	PTTG1	ADIRF	CD93	CCL5	CD24	MGP	TMEM158	SERPINB4	TIMP1
9	PLA2G2A	HOPX	CXCL11	COL1A1	PTGDS	LOX	KRT8	HMGN2	CCDC102B	ADGRL4	CXCR4	IGKC	GREM1	RSAD2	MT1X	SERPINB2
10	PI16	TSPAN8	TNFSF10	COL6A3	CLU	MT2A	TM4SF1	HIST1H4C	A2M	RBP7	CCL3	WFDC2	APOE	IGFBP2	G0S2	CXCL6
11	DPT	FHL2	STAT1	RGCC	CADM3	ERO1A	C3	PCLAF	ACTA2	CLEC14A	C1QC	GPRC5A	CD36	MMP13	MT1E	NAMPT
12	SFRP1	TCEAL4	MX2	ISLR	SELENOP	HSPA5	CALB2	BIRC5	COL18A1	CD36	FCER1G	C19orf33	IGF1	WNT5A	PI15	SERPINB4
13	WISP2	SPARCL1	PARP14	AEBP1	RARRES1	NAMPT	ITLN1	CCNB1	MEF2C	CD34	PTPRC	CD9	CYP1B1	SULF1	FTH1	IL6
14	SRPX	NR2F2	OAS1	COL6A1	CLDN1	BNIP3	PROCR	ASPM	HIGD1B	PCDH17	LYZ	KRT19	TSHZ2	BST2	SH3BGRL3	MT1G
15	SFRP2	ABLIM1	IFI6	SFRP4	IER2	GAPDH	CLDN1	TYMS	NOTCH3	RNASE1	GNLY	ELF3	C7	MTRNR2L8	FGF7	SOD2
16	IGFBP6	KIF5C	GBP1	COL10A1	HLA-DRA	ADM	PLAT	TPX2	CRIP1	ITGA6	LAPTM5	CLDN4	CHI3L1	IFI27	MT3	HP
17	FBLN2	STAR	BST2	COL5A2	LXN	FTH1	CCDC80	HMGB2	ITGA1	SRGN	CD14	MAL2	ADH1B	CTSK	OST4	CCDC71L
18	PLAC9	TRABD2A	RNF213	PALLD	EFEMP1	SLC2A1	SAA1	CDKN3	LHFPL6	CDH5	HLA-DQB1	MUC1	MTRNR2L12	CHN1	GAPDH	ALDH1A3
19	FBLN5	SYNPO2	DDX58	SFRP2	CENPW	HSP90B1	SLC39A8	HMGB1	C11orf96	FLT1	HLA-DQA1	TACSTD2	AC092069.1	TMEM176A	LDHA	CFB
20	FBN1	DKK3	LAP3	FN1	ATF3	NUPR1	AQP1	CKS2	ANGPT2	EMCN	AIF1	SMIM22	CYR61	TMEM176B	PLIN2	SLC39A8
21	ABCA8	CHL1	IFI27	MAFB	IER3	LDHA	TIMP1	SMC4	MYH11	MMRN1	NKG7	FOLR1	LRRC75A	PVALB	TGFBI	MT1X
22	PDK4	TCF21	IFIH1	INHBA	HLA-DRB1	PLOD2	CXCL6	UBE2C	EPS8	TM4SF18	CCL4L2	CP	POSTN	PRRX1	TPI1	CXCL8
23	MGST1	PRELP	MMP13	COL3A1	CYR61	CA9	S100A10	TUBB	UACA	PTPRB	FCGR3A	S100A1	ACTA2	ISG20	TMEM258	PRG4
24	FMO2	GATM	XAF1	HTRA1	TIMP1	IER3	HLA-DRB1	UBE2S	SPARCL1	RAMP3	CD52	CST6	CXCL12	NTM	CTSC	ABI3BP
25	ZFP36	SCN7A	EPSTI1	SEPT11	PIEZO2	CA12	SOD2	NUCKS1	PDGFA	SOX18	CD2	MSLN	GPX3	SERPINF1	TMA7	CA12
26	MT1A	ALKAL2	OASL	PLAU	PTGIS	ENO1	WFDC2	DLGAP5	NR2F2	CALCRL	LTB	MAL	CFD	AEBP1	UQCRQ	PLA2G2A
27	DCN	KLHDC8A	SAMD9	TCF4	ADAMTS1	DDIT4	PTGIS	TK1	GJC1	BTNL9	ITGB2	TCIM	WISP2	PSME1	EIF4EBP1	SBSN
28	GPC3	DIRAS3	IFI35	ANTXR1	SAT1	FAM162A	SAT1	KPNA2	ADGRF5	ESAM	CORO1A	BCAM	TAGLN	MME	ATP5ME	CLDN1
29	MFAP4	DHRS2	CCL11	PRRX1	SOD2	G0S2	RARRES1	TUBB4B	PLAC9	LMO2	HCST	CLDN7	SERPINF1	TCF4	PPP1R14B	DES
30	ABCA6	FOXL2	HERC5	LUM	DES	AL133453.1	SGK1	CENPE	CARMN	GIMAP7	GZMA	SCGB2A1	APOD	CD63	SERPINE1	PTX3
31	ABCA9	NRXN2	GBP4	COLEC12	WNT4	TGFBI	MT1G	TUBA1C	PDGFRB	NOTCH4	FYB1	ERP27	MALAT1	LUM	CADM3	CADM3
32	DEPP1	PALMD	OAS3	PMEPA1	EGFL6	EGLN3	RARRES3	JPT1	TINAGL1	SLCO2A1	CD69	SFTA2	NEAT1	NCAM2	ACTB	PDLIM1
33	SCARA5	LAMA2	EIF2AK2	SULF1	NFKBIZ	UPP1	CXCL1	KIF20B	COX4I2	MMRN2	ARHGDIB	MUC16	BGN	TMSB4X	C4orf48	PDZK1IP1
34	CXCL12	PLCXD3	MMP11	DIO2	RARRES3	PGK1	SELENOP	CKAP2	HES4	APLNR	SAMSN1	HMGA1	VMP1	MMP2	UBL5	CHI3L2
35	LHFPL6	WIPF3	LY6E	ITGB5	JUN	PTGS2	HLA-DRA	HMMR	TBX2	ICAM2	CD3D	S100A9	MMP2	ANTXR1	S100A11	CALB2
36	SVEP1	TMEM233	PLSCR1	FAP	METTL7A	ERRFI1	CEMIP	SMC2	CAV1	ECSCR	LCP1	TPI1	MBNL1	LINC01705	HSD11B1	PAPPA
37	PLPP3	SIGLEC11	CMPK2	CDH11	OSR1	CLEC2B	SERPINB2	NASP	STEAP4	ROBO4	MS4A6A	AGR2	POLR2J3-1	SOST	ENO1	VTN
38	PRELP	RASL11B	OAS2	MARCKS	C2	ID2	CFB	PCNA	MYL9	CXorf36	ALOX5AP	UCA1	ITPRIPL2	HTRA1	SERF2	OSR1
39	TXNIP	PRKAR2B	SAMD9L	MMP2	RHOU	HERPUD1	IL6ST	HELLS	GUCY1A2	NOSTRIN	CD53	UPK1B	EIF3L	HLA-C	TMSB10	EPGN
40	ITM2A	BMPR1B	TYMP	COL5A1	IL6ST	WSB1	DAB2	PBK	SEPT4	CYYR1	SLA	UCP2	SARS	TXN	COX7C	GRIA2
41	CD34	GREB1	ISG20	SPARC	SOD3	CXCL8	KRT7	RRM2	TAGLN	FAM167B	GZMB	SPINK1	ZCCHC24	THY1	SSR4	PI3
42	RASD1	PCSK6	IFI44	SDC1	TXNIP	SLC16A3	KLK11	MAD2L1	FAM162B	PCAT19	CD37	S100A14	LARP1	CTHRC1	ATP5F1E	CXADR
43	EFEMP1	ARX	TNFSF13B	TMEM158	PLS3	STC1	MAF	ANLN	TFPI	TIE1	CYTIP	CRIP1	SPPL2A	SELENOM	TREM1	PICSAR
44	FSTL1	P3H2	CTSK	NTM	CHI3L1	SOD2	CYB5A	CCNB2	JAG1	GIMAP4	CCL3L1	SCNN1A	COL6A3	HLA-A	PTPRN	LINC01133
45	COL14A1	CDHR1	PSMB9	LRRC17	NFKBIA	P4HB	SAA2	H2AFV	S100A4	MYCT1	CD163	TMEM238	EIF3A	WNT10A	CXCL5	UPK3B
46	TSHZ2	ECEL1	IL32	THY1	ZFP36	C15orf48	DES	CKS1B	PTP4A3	ESM1	MS4A7	VAMP8	ARID1B	LGALS1	IL33	FRAS1
47	SOCS3	HS3ST1	IDO1	MDK	FGF7	BNIP3L	CD74	DEK	PGF	RAPGEF5	CD48	ITGB8	PCSK7	HLA-B	SPOCD1	MARCH3
48	FBLN1	ARHGEF28	WARS	COL6A2	CD74	SERPINE2	EFEMP1	CDK1	ENPEP	CCL14	GPR183	FXYD3	PLTP	COL6A3	P2RY6	KLK11
49	LAMA2	ARHGAP25	SP100	GJA1	IL1R1	MANF	CLU	RAN	CD36	MCTP1	GMFG	SLC34A2	NFYB	PTMA	C5orf46	AADAC
50	CHRDL1	SERPINA5	TRIM22	EPYC	TNFAIP3	DDIT3	IGFBP6	DTYMK	AVPR1A	PREX2	GIMAP4	GSTP1	LUM	B2M	PTGS1	SERPINB3

2 Clusters

2.1 sizes

enframe(sort(table(seu_obj$cluster_label))) %>% 
  mutate(name = ordered(name, levels = rev(name))) %>% 
  ggplot() +
  geom_bar(aes(name, value), stat = "identity") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(y = c("#cells"), x = "cluster")

2.2 UMAP

alpha_lvl <- ifelse(nrow(plot_data) < 20000, 0.2, 0.1)
pt_size <- ifelse(nrow(plot_data) < 20000, 0.2, 0.05)

common_layers_disc <- list(  
  geom_point(size = pt_size, alpha = alpha_lvl),
  NoAxes(),
  guides(color = guide_legend(override.aes = list(size = 2, alpha = 1))),
  labs(color = "")
)

common_layers_cont <- list(  
  geom_point(size = pt_size, alpha = alpha_lvl),
  NoAxes(),
  scale_color_gradientn(colors = viridis(9)),
  guides(color = guide_colorbar())
)

ggplot(plot_data, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  #facet_wrap(~therapy) +
  ggtitle("Sub cluster")

3 Filtering out doublet clusters

3.1 UMAP

my_subtypes <- names(clrs$cluster_label[[coi]])
my_subtypes <- c(my_subtypes, unlist(lapply(paste0("_", 1:3), function(x) paste0(my_subtypes, x)))) %>% .[!str_detect(., "doublet|dissociated|Mito.high|OV\\.[0-9]")]
my_subtypes <- my_subtypes[my_subtypes %in% unique(seu_obj$cluster_label)]
my_subtypes <- my_subtypes[my_subtypes %in% names(clrs$cluster_label[[coi]])]

cells_to_keep <- colnames(seu_obj)[seu_obj$cluster_label %in% my_subtypes]
seu_obj_sub <- subset(seu_obj, cells = cells_to_keep)
seu_obj_sub <- RunUMAP(seu_obj_sub, dims = 1:50, reduction = "harmony", reduction.name = "umapharmony", reduction.key = "umapharmony_")
seu_obj_sub$cluster_label <- seu_obj$cluster_label[colnames(seu_obj) %in% colnames(seu_obj_sub)]
write_rds(seu_obj_sub, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_processed_filtered.rds"))
seu_obj_sub <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_processed_filtered.rds"))

plot_data_sub <- as_tibble(FetchData(seu_obj_sub, c(myfeatures, "cluster_label"))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cell_id = colnames(seu_obj_sub),
         cluster_label = ordered(cluster_label, levels = my_subtypes),
         )
  
if (cell_sort == "CD45+") {
plot_data_sub <- filter(plot_data_sub, sort_parameters != "singlet, live, CD45-", !is.na(tumor_supersite))
}

if (cell_sort == "CD45-") {
plot_data_sub <- filter(plot_data_sub, sort_parameters != "singlet, live, CD45+", !is.na(tumor_supersite))
}

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  ggtitle("Sub cluster") +
  #facet_wrap(~cluster_label) +
  scale_color_manual(values = clrs$cluster_label[[coi]])

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = patient_id_short)) + 
  common_layers_disc +
  ggtitle("Patient") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$patient_id_short)

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = tumor_supersite)) + 
  # geom_point(aes(umapharmony_1, umapharmony_2), 
  #            color = "grey90", size = 0.01, 
  #            data = select(plot_data_sub, -tumor_supersite)) +
  common_layers_disc +
  ggtitle("Site") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$tumor_supersite)

write_tsv(select(plot_data_sub, cell_id, everything(), -umapharmony_1, -umapharmony_2, -contains("RNA_")), paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_embedding.tsv"))

3.2 add module scores and pathway scores

signature_modules <- read_excel("_data/small/signatures/SPECTRUM v5 sub cluster markers.xlsx", sheet = 2, skip = 1, range = "M2:P100") %>%
  gather(module, gene) %>%
  na.omit() %>%
  group_by(module) %>%
  do(gene = c(.$gene)) %>%
  {setNames(.$gene, .$module)}

signature_modules$ISG.module <- c("CCL5", "CXCL10", "IFNA1", "IFNB1", "ISG15", "IFI27L2", "SAMD9L")

## compute expression module scores
for (i in 1:length(signature_modules)) {
  seu_obj_sub <- AddModuleScore(seu_obj_sub, features = signature_modules[i], name = names(signature_modules)[i])
  seu_obj_sub[[names(signature_modules)[i]]] <- seu_obj_sub[[paste0(names(signature_modules)[i], "1")]]
  seu_obj_sub[[paste0(names(signature_modules)[i], "1")]] <- NULL
  print(paste(names(signature_modules)[i], "DONE"))
}

## [1] "CD8.Cytotoxic DONE"
## [1] "CD8.Dysfunctional DONE"
## [1] "CD8.Naive DONE"
## [1] "CD8.Predysfunctional DONE"
## [1] "ISG.module DONE"

## compute progeny scores
progeny_list <- seu_obj_sub@assays$RNA@data[VariableFeatures(seu_obj_sub),] %>%
  as.matrix %>%
  progeny %>%
  as.data.frame %>%
  as.list

names(progeny_list) <- make.names(paste0(names(progeny_list), ".pathway"))

for (i in 1:length(progeny_list)) {
  seu_obj_sub <- AddMetaData(seu_obj_sub, metadata = progeny_list[[i]],
                             col.name = names(progeny_list)[i])
}

write_rds(seu_obj_sub, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_processed_filtered_annotated.rds"))
seu_obj_sub <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_processed_filtered_annotated.rds"))

3.3 marker heatmap

marker_top_tbl <- marker_sheet[,-1] %>% 
  slice(1:10) %>% 
  as.list %>% 
  .[!str_detect(names(.), "doublet|dissociated|Mito.high|OV\\.[0-9]")] %>% 
  enframe("cluster_label_x", "gene") %>% 
  unnest(gene)

plot_data_markers <- as_tibble(FetchData(seu_obj_sub, c("cluster_label", myfeatures, unique(marker_top_tbl$gene)))) %>% 
  gather(gene, value, -c(1:(length(myfeatures)+1))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cluster_label = ordered(cluster_label, levels = my_subtypes)) %>% 
  group_by(cluster_label, gene) %>% 
  summarise(value = mean(value, na.rm = T)) %>% 
  group_by(gene) %>% 
  mutate(value = scales::rescale(value)) %>% 
  left_join(marker_top_tbl, by = "gene") %>% 
  mutate(cluster_label_x = ordered(cluster_label_x, levels = rev(names(clrs$cluster_label[[coi]]))))

ggplot(plot_data_markers) +
  geom_tile(aes(gene, cluster_label, fill = value)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  facet_grid(~cluster_label_x, scales = "free", space = "free") +
  scale_fill_gradientn(colors = viridis(9)) +
  labs(fill = "Scaled\nexpression") +
  theme(aspect.ratio = 1,
        axis.line = element_blank(),
        axis.ticks = element_blank(),
        axis.title = element_blank())

# ggsave(paste0("_fig/002_marker_heatmap_", coi, ".pdf"), width = nrow(marker_top_tbl)/6, height = 5)

3.4 composition

3.4.1 per site

comp_site_tbl <- plot_data_sub %>%
  filter(!is.na(tumor_supersite)) %>% 
  group_by(cluster_label, tumor_supersite) %>%
  tally %>%
  group_by(tumor_supersite) %>%
  mutate(nrel = n/sum(n)*100) %>%
  ungroup

pnrel_site <- ggplot(comp_site_tbl) +
  geom_bar(aes(tumor_supersite, nrel, fill = cluster_label),
           stat = "identity", position = position_stack()) +
  scale_y_continuous(expand = c(0, 0)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(fill = "Cluster", y = "Fraction [%]", x = "") +
  scale_fill_manual(values = clrs$cluster_label[[coi]])

pnabs_site <- ggplot(comp_site_tbl) +
  geom_bar(aes(tumor_supersite, n, fill = cluster_label),
           stat = "identity", position = position_stack()) +
  scale_y_continuous(expand = c(0, 0)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(fill = "Cluster", y = "# cells", x = "") +
  scale_fill_manual(values = clrs$cluster_label[[coi]])

plot_grid(pnabs_site, pnrel_site, ncol = 2, align = "h")

# ggsave(paste0("_fig/02_deep_dive_", coi, "_comp_site.pdf"), width = 8, height = 4)

3.4.2 per sample

comp_tbl_sample_sort <- plot_data_sub %>% 
  group_by(tumor_subsite, tumor_supersite, patient_id, consensus_signature, therapy, cluster_label) %>% 
  tally %>% 
  group_by(tumor_subsite, tumor_supersite, patient_id, consensus_signature, therapy) %>% 
  mutate(nrel = n/sum(n)*100,
         nsum = sum(n),
         log10n = log10(n),
         label_supersite = "Site",
         label_mutsig = "Signature",
         label_therapy = "Rx") %>% 
  ungroup %>% 
  arrange(desc(therapy), tumor_supersite) %>% 
  mutate(tumor_subsite_rx = paste0(tumor_subsite, "_", therapy)) %>% 
  mutate(tumor_subsite = ordered(tumor_subsite, levels = unique(tumor_subsite)),
         tumor_subsite_rx = ordered(tumor_subsite_rx, levels = unique(tumor_subsite_rx))) %>% 
  arrange(patient_id) %>% 
  mutate(label_patient_id = ifelse(as.logical(as.numeric(fct_inorder(as.character(patient_id)))%%2), "Patient1", "Patient2"))

sample_id_x_tbl <- plot_data_sub %>% 
  mutate(sort_short_x = cell_sort) %>% 
  distinct(patient_id, sort_short_x, tumor_subsite, therapy, sample) %>% 
  unite(sample_id_x, patient_id, sort_short_x, tumor_subsite, therapy) %>% 
  arrange(sample_id_x)

comp_tbl_sample_sort %>% 
  mutate(sort_short_x = cell_sort) %>% 
  unite(sample_id_x, patient_id, sort_short_x, tumor_subsite_rx) %>% 
  select(sample_id_x, cluster_label, n, nrel, nsum) %>% 
  left_join(sample_id_x_tbl, by = "sample_id_x") %>% 
  write_tsv(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v7/", coi, "_subtype_compositions.tsv"))

ybreaks <- c(500, 1000, 2000, 4000, 6000, 8000, 10000, 15000, 20000)

max_cells_per_sample <- max(comp_tbl_sample_sort$nsum)
ymaxn <- ybreaks[max_cells_per_sample < ybreaks][1]

comp_plot_wrapper <- function(y = "nrel", switch = NULL) {
  if (y == "nrel") ylab <- paste0("Fraction\nof cells [%]")
  if (y == "n") ylab <- paste0("Number\nof cells")
  p <- ggplot(comp_tbl_sample_sort, 
              aes_string("tumor_subsite_rx", y, fill = "cluster_label")) + 
    facet_grid(~patient_id, space = "free", scales = "free", switch = switch) +
    coord_cartesian(clip = "off") + 
    scale_fill_manual(values = clrs$cluster_label[[coi]]) + 
    theme(axis.text.x = element_blank(),
          axis.title.y = element_text(angle = 0, vjust = 0.5, hjust = 0.5, 
                                      margin = margin(0, -0.4, 0, 0, unit = "npc")),
          axis.ticks.x = element_blank(),
          axis.title.x = element_blank(),
          axis.line.x = element_blank(),
          strip.text.y = element_blank(),
          strip.text.x = element_blank(),
          strip.background.y = element_blank(),
          strip.background.x = element_blank(),
          plot.margin = margin(0, 0, 0, 0, "npc")) + 
    labs(x = "",
         y = ylab) +
    guides(fill = FALSE)
  if (y == "nrel") p <- p + 
    geom_bar(stat = "identity") +
    scale_y_continuous(expand = c(0, 0), 
                       breaks = c(0, 50, 100), 
                       labels = c("0", "50", "100"))
  if (y == "n") p <- p + 
    geom_bar(stat = "identity", position = position_stack()) +
    scale_y_continuous(expand = c(0, 0), 
                       breaks = c(0, ymaxn/2, ymaxn),
                       limits = c(0, ymaxn),
                       labels = c("", ymaxn/2, ymaxn)) +
    expand_limits(y = c(0, ymaxn)) +
    theme(panel.grid.major.y = element_line(linetype = 1, color = "grey90", size = 0.5))
  return(p)
} 

common_label_layers <- list(
  geom_tile(color = "white", size = 0.15),
  theme(axis.text.x = element_blank(),
        axis.ticks = element_blank(),
        axis.title.x = element_blank(),
        axis.line.x = element_blank(),
        strip.text = element_blank(),
        strip.background = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "npc")),
  scale_y_discrete(expand = c(0, 0)),
  labs(y = ""),
  guides(fill = FALSE),
  facet_grid(~patient_id, 
             space = "free", scales = "free")
)

comp_label_site <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_supersite, patient_id), 
       aes(tumor_subsite_rx, label_supersite, 
           fill = tumor_supersite)) + 
  scale_fill_manual(values = clrs$tumor_supersite) +
  common_label_layers

comp_label_rx <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_therapy, consensus_signature, patient_id), 
       aes(tumor_subsite_rx, label_therapy, 
           fill = therapy)) + 
  scale_fill_manual(values = c(`post-Rx` = "gold3", `pre-Rx` = "steelblue")) +
  common_label_layers

comp_label_mutsig <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_mutsig, consensus_signature, patient_id), 
       aes(tumor_subsite_rx, label_mutsig, 
           fill = consensus_signature)) + 
  scale_fill_manual(values = clrs$consensus_signature) +
  common_label_layers

patient_label_tbl <- distinct(comp_tbl_sample_sort, patient_id, .keep_all = T)

comp_label_patient_id <- ggplot(patient_label_tbl, aes(tumor_subsite_rx, label_patient_id)) + 
  scale_fill_manual(values = clrs$consensus_signature) +
  geom_text(aes(tumor_subsite_rx, label_patient_id, label = patient_id)) +
  facet_grid(~patient_id, 
             space = "free", scales = "free") +
  coord_cartesian(clip = "off") + 
  theme_void() +
  theme(strip.text = element_blank(),
        strip.background = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "npc"))

hist_plot_wrapper <- function(x = "nrel") {
  if (x == "nrel") {
    xlab <- paste0("Fraction of cells [%]")
    bw <- 5
  }
  if (x == "log10n") {
    xlab <- paste0("Number of cells")
    bw <- 0.2
  }
  p <- ggplot(comp_tbl_sample_sort) +
    ggridges::geom_density_ridges(
      aes_string(x, "cluster_label", fill = "cluster_label"), color = "black",
      stat = "binline", binwidth = bw, scale = 3) +
    facet_grid(label_supersite~., 
               space = "free", scales = "free") +
    scale_fill_manual(values = clrs$cluster_label[[coi]]) + 
    theme(axis.text.y = element_blank(),
          axis.ticks.y = element_blank(),
          axis.title.y = element_blank(),
          axis.line.y = element_blank(),
          strip.text = element_blank(),
          strip.background = element_blank(),
          plot.margin = margin(0, 0, 0, 0, "npc")) +
    scale_x_continuous(expand = c(0, 0)) +
    scale_y_discrete(expand = c(0, 0)) +
    guides(fill = FALSE) +
    labs(x = xlab)
  if (x == "log10n") p <- p + expand_limits(x = c(0, 3)) + 
    scale_x_continuous(expand = c(0, 0), 
                       labels = function(x) parse(text = paste("10^", x)))
  return(p)
}

pcomp1 <- comp_plot_wrapper("n")
pcomp2 <- comp_plot_wrapper("nrel")

pcomp_grid <- plot_grid(comp_label_patient_id, 
                        pcomp1, pcomp2, 
                        comp_label_site, comp_label_rx, comp_label_mutsig,
                        ncol = 1, align = "v", 
                        rel_heights = c(0.15, 0.33, 0.33, 0.06, 0.06, 0.06))

phist1 <- hist_plot_wrapper("log10n")

pcomp_hist_grid <- ggdraw() +
  draw_plot(pcomp_grid, x = 0.01, y = 0, width = 0.85, height = 1) +
  draw_plot(phist1, x = 0.87, y = 0.05, width = 0.12, height = 0.8)

pcomp_hist_grid

# ggsave(paste0("_fig/02_composition_v7_",coi,".pdf"), pcomp_hist_grid, width = 10, height = 2)

3.4.3 site specific cluster enrichment

comp_tbl_z <- comp_tbl_sample_sort %>% 
  filter(therapy == "pre-Rx",
         !(tumor_supersite %in% c("Ascites", "Other"))) %>% 
  group_by(patient_id, cluster_label) %>% 
  arrange(patient_id, cluster_label, nrel) %>% 
  mutate(rank = row_number(nrel),
         z_rank = scales::rescale(rank)) %>% 
  mutate(mean_nrel = mean(nrel, na.rm = T),
         sd_nrel = sd(nrel, na.rm = T),
         z_nrel = (nrel - mean_nrel) / sd_nrel) %>% 
  ungroup()

ggplot(comp_tbl_z) +
  geom_boxplot(aes(tumor_supersite, z_nrel, color = tumor_supersite), 
               outlier.shape = NA) +
  geom_point(aes(tumor_supersite, z_nrel, color = tumor_supersite), position = position_jitter(), size = 0.1) +
  facet_grid(~cluster_label, scales = "free_x", space = "free_x") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5),
        aspect.ratio = 5) +
  scale_color_manual(values = clrs$tumor_supersite)

ggplot(comp_tbl_z) +
  geom_boxplot(aes(tumor_supersite, z_rank, color = tumor_supersite), 
               outlier.shape = NA) +
  geom_point(aes(tumor_supersite, z_rank, color = tumor_supersite), position = position_jitter(), size = 0.1) +
  facet_grid(~cluster_label, scales = "free_x", space = "free_x") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5),
        aspect.ratio = 5) +
  scale_color_manual(values = clrs$tumor_supersite)

4 session info

devtools::session_info()

## ─ Session info ───────────────────────────────────────────────────────────────
##  setting  value                       
##  version  R version 3.6.2 (2019-12-12)
##  os       Debian GNU/Linux 10 (buster)
##  system   x86_64, linux-gnu           
##  ui       X11                         
##  language (EN)                        
##  collate  en_US.UTF-8                 
##  ctype    en_US.UTF-8                 
##  tz       Etc/UTC                     
##  date     2021-03-09                  
## 
## ─ Packages ───────────────────────────────────────────────────────────────────
##  package              * version    date       lib
##  abind                  1.4-5      2016-07-21 [2]
##  ape                    5.3        2019-03-17 [2]
##  assertthat             0.2.1      2019-03-21 [2]
##  backports              1.1.10     2020-09-15 [1]
##  bibtex                 0.4.2.2    2020-01-02 [2]
##  Biobase                2.46.0     2019-10-29 [2]
##  BiocGenerics           0.32.0     2019-10-29 [2]
##  BiocParallel           1.20.1     2019-12-21 [2]
##  bitops                 1.0-6      2013-08-17 [2]
##  boot                   1.3-24     2019-12-20 [3]
##  broom                  0.7.2      2020-10-20 [1]
##  callr                  3.4.2      2020-02-12 [1]
##  car                    3.0-8      2020-05-21 [1]
##  carData                3.0-4      2020-05-22 [1]
##  caTools                1.17.1.4   2020-01-13 [2]
##  cellranger             1.1.0      2016-07-27 [2]
##  class                  7.3-15     2019-01-01 [3]
##  cli                    2.0.2      2020-02-28 [1]
##  cluster                2.1.0      2019-06-19 [3]
##  codetools              0.2-16     2018-12-24 [3]
##  colorblindr          * 0.1.0      2020-01-13 [2]
##  colorspace           * 1.4-2      2019-12-29 [2]
##  cowplot              * 1.0.0      2019-07-11 [2]
##  crayon                 1.3.4      2017-09-16 [1]
##  curl                   4.3        2019-12-02 [2]
##  data.table             1.12.8     2019-12-09 [2]
##  DBI                    1.1.0      2019-12-15 [2]
##  dbplyr                 2.0.0      2020-11-03 [1]
##  DelayedArray           0.12.2     2020-01-06 [2]
##  DEoptimR               1.0-8      2016-11-19 [1]
##  desc                   1.2.0      2018-05-01 [2]
##  destiny              * 3.0.1      2020-01-16 [1]
##  devtools               2.2.1      2019-09-24 [2]
##  digest                 0.6.25     2020-02-23 [1]
##  dplyr                * 1.0.2      2020-08-18 [1]
##  e1071                  1.7-3      2019-11-26 [1]
##  ellipsis               0.3.1      2020-05-15 [1]
##  evaluate               0.14       2019-05-28 [2]
##  fansi                  0.4.1      2020-01-08 [2]
##  farver                 2.0.3      2020-01-16 [1]
##  fitdistrplus           1.0-14     2019-01-23 [2]
##  forcats              * 0.5.0      2020-03-01 [1]
##  foreign                0.8-74     2019-12-26 [3]
##  formattable            0.2.0.1    2016-08-05 [1]
##  fs                     1.5.0      2020-07-31 [1]
##  future                 1.15.1     2019-11-25 [2]
##  future.apply           1.4.0      2020-01-07 [2]
##  gbRd                   0.4-11     2012-10-01 [2]
##  gdata                  2.18.0     2017-06-06 [2]
##  generics               0.0.2      2018-11-29 [2]
##  GenomeInfoDb           1.22.0     2019-10-29 [2]
##  GenomeInfoDbData       1.2.2      2020-01-14 [2]
##  GenomicRanges          1.38.0     2019-10-29 [2]
##  ggplot.multistats      1.0.0      2019-10-28 [1]
##  ggplot2              * 3.3.2      2020-06-19 [1]
##  ggrepel                0.8.1      2019-05-07 [2]
##  ggridges               0.5.2      2020-01-12 [2]
##  ggthemes               4.2.0      2019-05-13 [1]
##  globals                0.12.5     2019-12-07 [2]
##  glue                   1.3.2      2020-03-12 [1]
##  gplots                 3.0.1.2    2020-01-11 [2]
##  gridExtra              2.3        2017-09-09 [2]
##  gtable                 0.3.0      2019-03-25 [2]
##  gtools                 3.8.1      2018-06-26 [2]
##  haven                  2.3.1      2020-06-01 [1]
##  hexbin                 1.28.0     2019-11-11 [2]
##  hms                    0.5.3      2020-01-08 [1]
##  htmltools              0.5.1.1    2021-01-22 [1]
##  htmlwidgets            1.5.1      2019-10-08 [2]
##  httr                   1.4.2      2020-07-20 [1]
##  ica                    1.0-2      2018-05-24 [2]
##  igraph                 1.2.5      2020-03-19 [1]
##  IRanges                2.20.2     2020-01-13 [2]
##  irlba                  2.3.3      2019-02-05 [2]
##  jsonlite               1.7.1      2020-09-07 [1]
##  KernSmooth             2.23-16    2019-10-15 [3]
##  knitr                  1.26       2019-11-12 [2]
##  labeling               0.3        2014-08-23 [2]
##  laeken                 0.5.1      2020-02-05 [1]
##  lattice                0.20-38    2018-11-04 [3]
##  lazyeval               0.2.2      2019-03-15 [2]
##  leiden                 0.3.1      2019-07-23 [2]
##  lifecycle              0.2.0      2020-03-06 [1]
##  listenv                0.8.0      2019-12-05 [2]
##  lmtest                 0.9-37     2019-04-30 [2]
##  lsei                   1.2-0      2017-10-23 [2]
##  lubridate              1.7.9.2    2020-11-13 [1]
##  magrittr             * 2.0.1      2020-11-17 [1]
##  MASS                   7.3-51.5   2019-12-20 [3]
##  Matrix                 1.2-18     2019-11-27 [3]
##  matrixStats            0.56.0     2020-03-13 [1]
##  memoise                1.1.0      2017-04-21 [2]
##  metap                  1.2        2019-12-08 [2]
##  mnormt                 1.5-5      2016-10-15 [2]
##  modelr                 0.1.8      2020-05-19 [1]
##  multcomp               1.4-12     2020-01-10 [2]
##  multtest               2.42.0     2019-10-29 [2]
##  munsell                0.5.0      2018-06-12 [2]
##  mutoss                 0.1-12     2017-12-04 [2]
##  mvtnorm                1.0-12     2020-01-09 [2]
##  nlme                   3.1-143    2019-12-10 [3]
##  nnet                   7.3-12     2016-02-02 [3]
##  npsurv                 0.4-0      2017-10-14 [2]
##  numDeriv               2016.8-1.1 2019-06-06 [2]
##  openxlsx               4.1.5      2020-05-06 [1]
##  pbapply                1.4-2      2019-08-31 [2]
##  pcaMethods             1.78.0     2019-10-29 [2]
##  pillar                 1.4.6      2020-07-10 [1]
##  pkgbuild               1.0.6      2019-10-09 [2]
##  pkgconfig              2.0.3      2019-09-22 [1]
##  pkgload                1.0.2      2018-10-29 [2]
##  plotly                 4.9.1      2019-11-07 [2]
##  plotrix                3.7-7      2019-12-05 [2]
##  plyr                   1.8.5      2019-12-10 [2]
##  png                    0.1-7      2013-12-03 [2]
##  prettyunits            1.1.1      2020-01-24 [1]
##  processx               3.4.2      2020-02-09 [1]
##  progeny              * 1.11.3     2020-10-22 [1]
##  proxy                  0.4-24     2020-04-25 [1]
##  ps                     1.3.2      2020-02-13 [1]
##  purrr                * 0.3.4      2020-04-17 [1]
##  R.methodsS3            1.7.1      2016-02-16 [2]
##  R.oo                   1.23.0     2019-11-03 [2]
##  R.utils                2.9.2      2019-12-08 [2]
##  R6                     2.4.1      2019-11-12 [1]
##  ranger                 0.12.1     2020-01-10 [1]
##  RANN                   2.6.1      2019-01-08 [2]
##  rappdirs               0.3.1      2016-03-28 [2]
##  RColorBrewer           1.1-2      2014-12-07 [2]
##  Rcpp                   1.0.4      2020-03-17 [1]
##  RcppAnnoy              0.0.16     2020-03-08 [1]
##  RcppEigen              0.3.3.7.0  2019-11-16 [2]
##  RcppHNSW               0.2.0      2019-09-20 [2]
##  RcppParallel           4.4.4      2019-09-27 [2]
##  RCurl                  1.98-1.1   2020-01-19 [1]
##  Rdpack                 0.11-1     2019-12-14 [2]
##  readr                * 1.4.0      2020-10-05 [1]
##  readxl               * 1.3.1      2019-03-13 [2]
##  rematch                1.0.1      2016-04-21 [2]
##  remotes                2.1.0      2019-06-24 [2]
##  reprex                 0.3.0      2019-05-16 [2]
##  reshape2               1.4.3      2017-12-11 [2]
##  reticulate             1.14       2019-12-17 [2]
##  rio                    0.5.16     2018-11-26 [1]
##  rlang                  0.4.8      2020-10-08 [1]
##  rmarkdown              2.0        2019-12-12 [2]
##  robustbase             0.93-6     2020-03-23 [1]
##  ROCR                   1.0-7      2015-03-26 [2]
##  rprojroot              1.3-2      2018-01-03 [2]
##  RSpectra               0.16-0     2019-12-01 [2]
##  rstudioapi             0.11       2020-02-07 [1]
##  rsvd                   1.0.3      2020-02-17 [1]
##  Rtsne                  0.15       2018-11-10 [2]
##  rvest                  0.3.6      2020-07-25 [1]
##  S4Vectors              0.24.2     2020-01-13 [2]
##  sandwich               2.5-1      2019-04-06 [2]
##  scales                 1.1.0      2019-11-18 [2]
##  scatterplot3d          0.3-41     2018-03-14 [1]
##  sctransform            0.2.1      2019-12-17 [2]
##  SDMTools               1.1-221.2  2019-11-30 [2]
##  sessioninfo            1.1.1      2018-11-05 [2]
##  Seurat               * 3.1.2      2019-12-12 [2]
##  SingleCellExperiment   1.8.0      2019-10-29 [2]
##  smoother               1.1        2015-04-16 [1]
##  sn                     1.5-4      2019-05-14 [2]
##  sp                     1.4-2      2020-05-20 [1]
##  stringi                1.5.3      2020-09-09 [1]
##  stringr              * 1.4.0      2019-02-10 [1]
##  SummarizedExperiment   1.16.1     2019-12-19 [2]
##  survival               3.1-8      2019-12-03 [3]
##  testthat               2.3.2      2020-03-02 [1]
##  TFisher                0.2.0      2018-03-21 [2]
##  TH.data                1.0-10     2019-01-21 [2]
##  tibble               * 3.0.4      2020-10-12 [1]
##  tidyr                * 1.1.2      2020-08-27 [1]
##  tidyselect             1.1.0      2020-05-11 [1]
##  tidyverse            * 1.3.0      2019-11-21 [2]
##  tsne                   0.1-3      2016-07-15 [2]
##  TTR                    0.23-6     2019-12-15 [1]
##  usethis                1.5.1      2019-07-04 [2]
##  uwot                   0.1.5      2019-12-04 [2]
##  vcd                    1.4-7      2020-04-02 [1]
##  vctrs                  0.3.5      2020-11-17 [1]
##  VIM                    6.0.0      2020-05-08 [1]
##  viridis              * 0.5.1      2018-03-29 [2]
##  viridisLite          * 0.3.0      2018-02-01 [2]
##  withr                  2.3.0      2020-09-22 [1]
##  xfun                   0.12       2020-01-13 [2]
##  xml2                   1.3.2      2020-04-23 [1]
##  xts                    0.12-0     2020-01-19 [1]
##  XVector                0.26.0     2019-10-29 [2]
##  yaml                   2.2.1      2020-02-01 [1]
##  zip                    2.0.4      2019-09-01 [1]
##  zlibbioc               1.32.0     2019-10-29 [2]
##  zoo                    1.8-7      2020-01-10 [2]
##  source                                 
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  Bioconductor                           
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Github (clauswilke/colorblindr@1ac3d4d)
##  R-Forge (R 3.6.2)                      
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  Bioconductor                           
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Github (saezlab/progeny@94bea1c)       
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.3)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
## 
## [1] /home/uhlitzf/R/lib
## [2] /usr/local/lib/R/site-library
## [3] /usr/local/lib/R/library

MSK SPECTRUM freeze major subset deep dives

Florian Uhlitz

2021-03-09

1 Cluster markers

1.1 Major Fibroblast.super markers for cell assign

1.2 Subtype currated markers

1.3 Subtype cluster markers

2 Clusters

2.1 sizes

2.2 UMAP

3 Filtering out doublet clusters

3.1 UMAP

3.2 add module scores and pathway scores

3.3 marker heatmap

3.4 composition

3.4.1 per site

3.4.2 per sample

3.4.3 site specific cluster enrichment

4 session info